Work–family interference in urban China: gender discrimination and the effects of work–family balance policies

The study investigates family responsibility discrimination in urban China, focusing on whether men or women face greater discrimination from observers when managing caregiving responsibilities that interfere with work (FIW) or when work interferes with family (WIF). Grounded in social role theory, which posits that societal gender role beliefs shape expectations (men as agentic breadwinners, women as communal caregivers), the authors note that while women traditionally face workplace discrimination, both genders can be sanctioned when acting counter to role expectations. Given persistent traditional beliefs in China alongside evolving norms, the research asks: In urban China, who is more likely to be discriminated against by observers in FIW and WIF contexts—men or women? And can firm work–family balance (WFB) policies mitigate such discrimination? The study centers on supervisors and irrelevant observers (general others or community neighbors) due to their potential to influence workers’ outcomes, particularly within a collectivist culture where social evaluations matter.

Prior research documents women’s greater exposure to workplace discrimination, including pay gaps and promotion barriers, and identifies family responsibility discrimination predominantly affecting women due to caregiving roles. Studies have often emphasized outcomes like work overload, stress, and work–family conflict and have largely examined Western contexts, leaving a gap regarding China. Social role theory explains that gendered expectations guide evaluations; deviations from role-congruent behavior elicit negative judgments. While Chinese gender role attitudes are traditionally conservative, rapid social changes and the integration of Eastern and Western norms may be altering expectations. Limited Chinese studies have considered discrimination against married women and mothers, but there is scarce evidence on how observers judge men versus women who shoulder family responsibilities, particularly under FIW and WIF. This study extends the literature by testing gender-differentiated observer discrimination in China and assessing whether organizational WFB policies buffer such biases.

Design and timeline: Four preregistered main experiments were conducted from late November 2021 to early June 2022 (N = 2577), complemented by five post-hoc/further tests in late August 2023 (N = 931). All studies used single-factor between-subjects designs with random assignment. Recruitment and sampling: Participants were recruited via Sojump, a large Chinese online crowdsourcing platform. Eligibility included age ≥18 and working or living in urban areas; supervisor experiments required managerial experience. The platform prevented repeat participation; captcha and attention checks ensured data quality. Compensation was ¥7–10 per participant. Observers: Key observer roles were supervisors (whose judgments emphasize work outcomes) and irrelevant observers (general others or community neighbors), reflecting collectivist cultural salience of social evaluation. Manipulations: - Gender (Experiments 1 and 2): Worker gender (male vs. female) communicated via text and pictures. - WFB policies (Experiments 3 and 4): High vs. low firm WFB policy scenarios based on the EU Work–Life Balance Directive and CSR for WLB. Contexts: - FIW (family interference in work): Worker often leaves work for family reasons (Experiments 1 and 3). - WIF (work interference in family): Teleworking spouse/parent neglects childcare while working at home (Experiments 2 and 4). Post-hoc/further tests varied responsibilities (care of aged parents), observer perspective (community neighbors; swapping supervisor vs. irrelevant observer perspectives), and analytic approach (OLS with controls). Measures: - Primary outcome: Contempt toward the worker, measured using Romani, Grappi, and Bagozzi’s contempt scale. - Mechanism measure (Study 2): Family conscientiousness rating using an established scale (ref. 48). - Manipulation checks: Perceived WFB policy strength (Studies 3 and 4). Controls for OLS: respondent gender, age, education, income, subjective SES, birthplace (urban vs. non-urban), regional pandemic risk, and industry gender composition. Procedures: Participants read the scenario corresponding to their assigned condition and rated contempt; demographics and (where applicable) manipulation check items followed. Statistical analysis: Primary analyses used two-sided independent-samples t-tests (95% CIs, effect size r computed as r = t^2/(t^2 + df)). Robustness checks used OLS regressions with controls; subgroup analyses split samples by subjective SES, income, education, industry dominance, and birthplace; differences assessed via seemingly unrelated estimation (Suest). Sample sizes and composition: - Experiment 1 (FIW, supervisors): N = 650 (325 male-target, 325 female-target). - Experiment 2 (WIF, irrelevant observers): N = 816 (403 male-target, 413 female-target). - Experiment 3 (FIW, supervisors, WFB high vs. low): N = 513 (255 high, 258 low). - Experiment 4 (WIF, irrelevant observers, WFB high vs. low): N = 598 (299 high, 299 low). Post-hoc/further tests: - Post-hoc (WIF, community neighbors): N = 132. - 1i (FIW, supervisors, care of aged parents): N = 212. - 2i (WIF, irrelevant observers, care of aged parents): N = 196. - 1ii (FIW, irrelevant observers): N = 211. - 2ii (WIF, supervisors): N = 180. Manipulation check results: Perceived WFB was significantly higher in high-policy conditions (Study 3: M = 5.77 vs. 2.85, t = −32.42, p < 0.001; Study 4: M = 5.89 vs. 3.05, t = −30.19, p < 0.001).

Main experiments: - Study 1 (FIW, supervisors): Male workers elicited more supervisor contempt than female workers (Mdiff = 0.387, se = 0.110, p < 0.001; Male M = 3.352, SD = 1.408; Female M = 2.965, SD = 1.379; r = 0.138). - Study 2 (WIF, irrelevant observers): Male workers elicited more contempt than female workers (Mdiff = 0.232, se = 0.081, p < 0.01; Male M = 2.487, SD = 1.180; Female M = 2.255, SD = 1.128; r = 0.100). Mechanism (Study 2): Males were rated lower on family conscientiousness than females (Mdiff = −0.335, se = 0.069, p < 0.001). - Study 3 (FIW, supervisors, WFB policies): High WFB policies reduced contempt toward male workers relative to low WFB (Mdiff = −0.327, se = 0.125, p < 0.01; High M = 3.054, SD = 1.365; Low M = 3.381, SD = 1.448; r = 0.116). - Study 4 (WIF, irrelevant observers, WFB policies): No significant difference between high and low WFB conditions (Mdiff = −0.042, se = 0.106, n.s.; High M = 2.591, SD = 1.232; Low M = 2.633, SD = 1.363; r = 0.016). Subgroup analyses (selected): - Lower subjective SES participants showed significantly more contempt for males vs. females in both contexts: FIW Mdiff = 0.397 (se = 0.112, p < 0.001); WIF Mdiff = 0.244 (se = 0.083, p < 0.01). - FIW: Higher and lower income participants showed more contempt for men (higher income Mdiff = 0.477, se = 0.239, p < 0.05; lower income Mdiff = 0.358, se = 0.123, p < 0.01). WIF: only lower income showed more contempt for men (Mdiff = 0.236, se = 0.085, p < 0.01). - Higher education participants showed more contempt for men in both contexts (FIW Mdiff = 0.453, se = 0.120, p < 0.001; WIF Mdiff = 0.297, se = 0.093, p < 0.001). - Male-dominated industries: stronger contempt for males (FIW Mdiff = 0.542, se = 0.132, p < 0.001; WIF Mdiff = 0.307, se = 0.101, p < 0.01). - Urban- and rural-born participants both showed more contempt for males in FIW (urban Mdiff = 0.380, se = 0.134, p < 0.01; rural Mdiff = 0.401, se = 0.183, p < 0.05) and WIF (urban Mdiff = 0.266, se = 0.114, p < 0.01; rural Mdiff = 0.187, se = 0.113, p < 0.10). WFB policy effects across groups (Study 3, FIW): Significant or marginal reductions in contempt under high WFB for higher subjective SES (Mdiff = −1.096, se = 0.363, p < 0.01), higher education (Mdiff = −0.340, se = 0.132, p < 0.01), higher income (Mdiff = −0.469, se = 0.246, p < 0.10), lower income (Mdiff = −0.271, se = 0.143, p < 0.10), male-dominated industries (Mdiff = −0.365, se = 0.143, p < 0.05), urban-born (Mdiff = −0.318, se = 0.154, p < 0.05), and rural-born (Mdiff = −0.353, se = 0.202, p < 0.10). No WFB mitigation effects for WIF (Study 4) across subgroups. Post-hoc and further tests: - Community neighbors (WIF): more contempt for men (Mdiff = 0.369, se = 0.216, p < 0.05). - Care of aged parents scenarios replicated male-targeted discrimination in FIW (Mdiff = 0.433, se = 0.198, p < 0.05) and WIF (Mdiff = 0.345, se = 0.188, p < 0.05). - Perspective switch: Irrelevant observers in FIW showed marginally more contempt for men (Mdiff = 0.290, se = 0.199, p < 0.10); supervisors in WIF showed a nonsignificant tendency toward less contempt for men (Mdiff = −0.115, se = 0.206, n.s.). Robustness: OLS with controls confirmed main results: gender effect significant in Studies 1, 2, and 3; WFB effect nonsignificant in Study 4; controls behaved plausibly (e.g., regional risk positive).

The findings address the core question by showing that, contrary to conventional wisdom that women face greater workplace discrimination, in urban China men garner more contempt from observers when family responsibilities interfere with work (FIW) and when work interferes with family (WIF). This aligns with role incongruity for men in caregiving and suggests rising expectations for men’s participation in family responsibilities, particularly salient during telework. Supervisors, focused on work outcomes, penalize male FIW, but their contempt is mitigated when firms signal strong WFB policies and a family-friendly culture. Irrelevant observers (society/community), who emphasize family role fulfillment, show heightened contempt for men in WIF, and this is not attenuated by organizational WFB policies, indicating limited reach of firm policies into societal evaluations. Subgroup patterns indicate stronger traditional gender role endorsements among those with lower subjective SES and in male-dominated industries, amplifying contempt toward men who deviate from breadwinner norms. The results broaden understanding of gender discrimination in a non-Western context, revealing that both gender norms and evolving expectations shape observer judgments, and that policy levers can be effective in some contexts (FIW within firms) but not others (WIF in family/social domains).

This study demonstrates that in urban China, men face greater observer discrimination than women when managing family responsibilities, both when family interferes with work and when work interferes with family. Firm-level WFB policies effectively reduce supervisors’ contempt toward men in FIW scenarios, but do not mitigate broader societal contempt toward men in WIF scenarios. Contributions include: reframing family responsibility discrimination as potentially targeting men under certain role incongruities; highlighting the differential effectiveness of organizational policies depending on observer type and context; and offering nuanced, non-Western evidence on gendered evaluations. Future research should identify additional interventions beyond firm WFB policies to alleviate discrimination in family/social contexts, examine the long-term impact of widespread telework on gender norms and caregiving divisions, and explore other observers and institutional mechanisms that may shape or buffer discriminatory judgments.